REVIEW 7 minor 15 references
On the Gromov-Hausdorff distance between the cloud of bounded metric spaces and a cloud with nontrivial stabilizer
T0 review · 0 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read This paper proves that the cloud of bounded metric spaces and the cloud containing the real line are separated by infinite Gromov–Hausdorff distance.
desk verdict The main theorem is real and the proof mostly works, but the abstract contradicts it and the exposition needs a serious cleanup before publication. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
Four ingredients carry the argument. (1) Clouds are proper-class equivalence classes of metric spaces under finite Gromov–Hausdorff distance; the paper proves every cloud is a proper class. (2) The stabilizer $\mathrm{St}([X])$ is the multiplicative group of positive scalings $\lambda$ with $[X] = [\lambda X]$, and a cloud with nontrivial stabilizer has a unique center, the space $M$ with $M = \lambda M$ for every $\lambda$ in the stabilizer. (3) The center-image theorem (Theorem 4.3) says that under a correspondence of finite distortion $\varepsilon$ between $[\Delta_1]$ and such a centered cloud, every image of $\Delta_1$ lies within $2\varepsilon$ of the center. (4) Inside $[\Delta_1]$, the Gromov–Hausdorff distance obeys the ultrametric inequality $|X_1,X_2| \leq \max\{|X_1,\Delta_1|, |X_2,\Delta_1|\}$, which converts the existence of $Y_1,Y_2$ with $|Y_1,Y_2| > r$ into the positive lower bound $\varepsilon \geq cr/4$.
What would settle it
Construct an explicit correspondence between $[\Delta_1]$ and $[R]$ with finite distortion, or a sequence of correspondences with distortions tending to $0$; the theorem predicts none exists, so an explicit construction would refute Corollary 6.3. A weaker check: find any cloud $[Z]$ satisfying the two inequalities of Theorem 6.2 for which a finite-distortion correspondence to $[\Delta_1]$ can be written down.
Extended reading notes
Core claim
The central claim is Theorem 6.2: if $[Z]$ is a cloud with a nontrivial stabilizer, $Z$ is its unique center, and there are spaces $Y_1, Y_2 \in [Z]$ with $\max\{|Y_1,Z|, |Y_2,Z|\} = r > 0$ and $|Y_1,Y_2| > r$, then the cloud distance $d_{\mathrm{GH}}([\Delta_1],[Z])$ is infinite. The proof first notes that clouds sharing a nontrivial stabilizer element have distance either $0$ or $\infty$ (Lemma 6.1), then rules out $0$ by showing any correspondence with distortion $\varepsilon$ forces $\varepsilon \geq c r / 4$ for some $c > 0$. Corollary 6.3 applies this to $[R]$: with $Y_1 = R$ and $Y_2 = \widetilde{R}$ (the real line with one extra point at unit $L^1$-distance), Theorem 5.2 supplies $r = 1/2$ and $|Y_1,Y_2| > r$, so $d_{\mathrm{GH}}([\Delta_1],[R]) = \infty$.
Load-bearing premise
The load-bearing premise is that any two proper classes admit a bijection, so a correspondence between any two clouds always exists and the cloud distance is defined; this is a global-choice axiom that the paper neither states nor proves.
Editorial extensions
If this is right
- The cloud of bounded metric spaces and the cloud containing the real line are infinitely far apart: no correspondence between them has finite distortion.
- Every cloud with a nontrivial stabilizer, a center, and a pair of spaces satisfying the two inequalities of Theorem 6.2 is also at infinite distance from the bounded cloud.
- The ultrametric inequality that holds in the bounded cloud fails in the real-line cloud, since $Z$ and $\widetilde{R}$ are each within $1/2$ of $R$ but more than $1/2$ from each other.
- The distance between any two clouds whose stabilizers intersect nontrivially is always either $0$ or $\infty$.
Reading between the lines
- Editorial inference: the infinite separation between $[\Delta_1]$ and $[R]$ suggests a general divide between the bounded-metric cloud and clouds whose centers are non-compact or have more than one end; such clouds may form a hierarchy ordered by infinite distance.
- Editorial inference: because $[\Delta_1]$ contains all compact metric spaces up to isometry, the result implies that compact geometry is not a dense approximation target for unbounded geometry in the cloud-level Gromov–Hausdorff distance, which would matter for shape-analysis pipelines that truncate unbounded data.
- Editorial inference: a natural testable extension is that $[\mathbb{R}^n]$ lies at infinite distance from $[\Delta_1]$ for every $n$, with the same two-spaces construction adapted to $\mathbb{R}^n$.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper works in the Gromov–Hausdorff class GH_0 of metric spaces up to zero distance and considers clouds, i.e., equivalence classes under finite Gromov–Hausdorff distance. It proves that every cloud is a proper class (Theorem 3.3), defines a Gromov–Hausdorff distance between clouds (Definition 3.5), and studies the multiplicative stabilizer of a cloud. The main result, Theorem 6.2, states that if a cloud [Z] has a nontrivial stabilizer with center Z and contains two spaces Y1,Y2 satisfying max(|Y1,Z|,|Y2,Z|)=r>0 and |Y1,Y2|>r, then the cloud distance between [Δ1] (the cloud of bounded metric spaces) and [Z] is infinite. Corollary 6.3 applies this to the real line, concluding that the distance between the cloud of bounded metric spaces and the cloud containing R is infinite. The proof combines a center-image theorem (Theorem 4.3), a scaling dichotomy (Lemma 6.1), and an example based on the integers and the real line with an extra point (Theorem 5.2).
Significance. The paper makes a genuine contribution to the geometry of the Gromov–Hausdorff class beyond the compact setting. The main theorem is surprising and nontrivial: it shows that the cloud of bounded metric spaces lies at infinite Gromov–Hausdorff distance from certain clouds with nontrivial stabilizers, including the real line. The proof is structurally coherent: Lemma 6.1 reduces the distance to 0 or ∞, Theorem 4.3 controls the image of the one-point space in a bounded-distortion correspondence, and the ultrametric inequality in the bounded cloud gives the final positive lower bound on distortion. The construction in Theorem 5.2 is explicit and is used directly in the main corollary. The paper contains no empirical fitting and no hidden parameters; the derivations are self-contained apart from standard references to prior work on clouds and centers.
minor comments (7)
- [Abstract and Section 1] The abstract states that "the distance between the cloud of bounded metric spaces and a cloud with a nontrivial stabilizer is finite", but Theorem 6.2 and Corollary 6.3 prove that this distance is infinite in the situations considered. This direct contradiction with the paper's main theorem must be corrected in revision.
- [Section 5, Theorem 5.2] The space Z is used in Theorem 5.2 and Corollary 6.3 but is never defined. It is apparently the set of integers with the usual metric, but this should be stated explicitly. In the proof, the symbol R is used both for the real line and for the correspondence R, which is confusing and should be disambiguated.
- [Lemma 6.1] The proof of Lemma 6.1 asserts without proof the scaling identity |λ[X], λ[Y]| = λ |[X], [Y]| for clouds. Since this identity is the basis of the "0 or ∞" dichotomy and hence is load-bearing for Theorem 6.2, please supply a short verification from Definition 3.5 or a precise reference.
- [Theorem 5.2(2)] The proof of the lower bound |Z, eR| > 1/2 is quite terse. In particular, the definition of the set N ("whose images lie in ...") and the application of Lemma 5.1 to obtain the estimate involving 1−ε should be expanded so that the argument can be followed without reconstructing it.
- [Corollary 6.3] The corollary says the conditions of Theorem 6.2 hold "with r = 1/2", but Theorem 5.2 gives only |Z,R| ≤ 1/2 and |eR,R| ≤ 1/2, not equality of the maximum to 1/2. The proof should instead take r = max(|Z,R|,|eR,R|), which is ≤ 1/2 and satisfies r < |Z,eR|, as needed.
- [Remark 3.4] The claim that a bijection exists between any two proper classes relies on the axiom of global choice in NBG set theory. The paper should state explicitly that it works in NBG with global choice, or provide a justification within the chosen framework.
- [Throughout] There are numerous typos and misprints, including "Intodiction" (heading), "non-utlrametric unequality" (Section 5 heading), "betweem" (Section 6 heading), "ba an element", and "posesses". Also, in the proof of Theorem 4.3 the symbol "∆" appears where "∆1" is meant. These should be corrected in a final editing pass.
Circularity Check
No significant circularity: the central cloud-distance theorem is derived from in-paper lemmas and standard prior structural results, not from its own conclusion.
full rationale
The central claim (Theorem 6.2 and Corollary 6.3) is not obtained by assuming what it proves. Theorem 6.2 combines Lemma 6.1, Theorem 4.3, and the ultrametric bound of Remark 2.12 to force a positive lower bound on the distortion of any correspondence between [Δ1] and [Z]; each of these inputs is either proved in the paper or is a standard, non-equivalent auxiliary fact. Theorem 4.3 is proved directly from the distortion inequalities of correspondences, and Lemma 6.1 follows from the scaling identity for cloud distances. The cited structural facts — centers of clouds from [12], stabilizer examples from [7,14], completeness from [7] — are supporting background, not renamed versions of the target distance computation. No parameter is fitted to a subset of data and then relabeled as a prediction. The abstract's statement that the distance is finite contradicts Corollary 6.3's infinite value, but that is an editing inconsistency, not a circular reduction. Remark 3.4's reliance on a bijection between proper classes is an axiom-level set-theoretic assumption (global choice), not a self-referential derivation. In the stated NBG framework, no circular step can be exhibited from the paper's own equations or definitions.
Assumptions & free parameters
assumptions (4)
- domain assumption Any two proper classes admit a bijection (global choice in NBG).
- standard math Every set of cardinal numbers has an upper bound (Lemma 3.1, cited from [15]).
- domain assumption Every cloud with nontrivial stabilizer has a unique center (Lemma 2.9, cited from [12]).
- standard math The ultrametric-type inequality in the cloud of bounded spaces: |X1,X2| <= max(|X1,Delta1|, |X2,Delta1|) (Remark 2.12).
Cite this review
Pith. "Pith review of On the Gromov-Hausdorff distance between the cloud of bounded metric spaces and a cloud with nontrivial stabilizer." pith.science (2026). https://pith.science/paper/5UG2TEAI
@misc{pith2026250523563,
author = {Pith},
title = {Pith review of: On the Gromov-Hausdorff distance between the cloud of bounded metric spaces and a cloud with nontrivial stabilizer},
year = {2026},
howpublished = {\url{https://pith.science/paper/5UG2TEAI}},
note = {Machine review of arXiv:2505.23563}
}
read the original abstract
The paper studies the class of all metric spaces considered up to zero Gromov-Hausdorff distance between them. In this class, we examine clouds - classes of spaces situated at finite Gromov-Hausdorff distances from a reference space. We prove that all clouds are proper classes. The Gromov-Hausdorff distance is defined for clouds similarly with the case of that for metric spaces. A multiplicative group of transformations of clouds is defined which is called stabilizer. We show that under certain restrictions the distance between the cloud of bounded metric spaces and a cloud with a nontrivial stabilizer is finite. In particular, the distance between the cloud of bounded metric spaces and the cloud containing the real line is calculated.
Reference graph
Works this paper leans on
-
[1]
Edwards, The Structure of Superspace
D. Edwards, The Structure of Superspace. In: Studies in Topology , ed. by Stavrakas N.M. and Allen K.R., New York, London, San Francisco, Academic Press, 1975
work page 1975
-
[2]
M. Gromov, Structures m´ etriques pour les vari´ et´ es riemanniennes, edited by Lafontaine and Pierre Pansu, 1981
work page 1981
-
[3]
M. Gromov, Metric structures for Riemannian and non-Riemannian spaces , Birkh¨ auser, ISBN 0-8176-3898-9 (translation with additional content), 1999
work page 1999
-
[4]
F. M´ emoli and S. Guillermo,Comparing point clouds , In Proceedings of the 2004 Eurograph- ics/ACM SIGGRAPH symposium on Geometry processing (SGP ’04). Association for Comput- ing Machinery, New York, NY, USA, 32-40. https://doi.org/10.1145/1057432.1057436, 2004
-
[5]
Multi-query Robotic Manipulator Task Sequencing with Gromov-Hausdorff Approximations
F. Sukkar, J. Wakulicz, K. M. B. Lee, W. Zhi and R. Fitch, Multi-query Robotic Manipulator Task Sequencing with Gromov–Hausdorff Approximations , ArXiv e-prints, arXiv:2209.04800, 2024. References 10
work page Pith review arXiv 2024
-
[6]
F. M´ emoli,Gromov–Hausdorff distances in Euclidean spaces, IEEE Computer Society Confer- ence on Computer Vision and Pattern Recognition Workshops, Anchorage, AK, USA, pp. 1-8, 2008
work page 2008
-
[7]
S. A. Bogatyy and A. A. Tuzhilin, Gromov–Hausdorff class: its completeness and cloud geom- etry, ArXiv e-prints, arXiv:2110.06101, 2021
arXiv 2021
-
[8]
J. von Neumann, Eine Axiomatisierung der Mengenlehre , Journal f¨ ur die Reine und Ange- wandte Mathematik, 1925
work page 1925
Show all 15 references
-
[9]
Bernays, A System of Axiomatic Set Theory–Part I , The Journal of Symbolic Logic, doi:10.2307/2268862, JSTOR 2268862, 1937
P. Bernays, A System of Axiomatic Set Theory–Part I , The Journal of Symbolic Logic, doi:10.2307/2268862, JSTOR 2268862, 1937
1937 doi
-
[10]
K. G¨ odel,The Consistency of the Axiom of Choice and of the Generalized Continuum Hypoth- esis with the Axioms of Set Theory (Revised ed.) , Princeton University Press, ISBN 978-0-691- 07927-1, 1940
1940
-
[11]
S. I. Borzov, A. O. Ivanov and A. A. Tuzhilin, Extendability of Metric Segments in Gromov– Hausdorff Distance, ArXiv e-prints, arXiv:2009.00458, 2020
2009 arXiv
-
[12]
S. I. Bogataya, S. A. Bogatyy, S. A., Redkozubov, V. V. & Tuzhilin A. A. , Clouds in Gromov– Hausdorff Class: their completeness and centers, ArXiv e-prints, arXiv:2202.07337, 2022
2022 arXiv
-
[13]
Burago, Y
D. Burago, Y. D. Burago, and S. O. Ivanov, A Course in Metric Geometry , American Mathe- matical Soc., 2001
2001
-
[14]
S. I. Bogataya and S. A. Bogatyy, Isometric Cloud Stabilizer , Topology and its Applications, Volume 329, 2023
2023
-
[15]
Levy, Basic set theory , Perspectives in mathematical logic, Springer-Verlag, Berlin, Heidel- berg, and New York, 1979
A. Levy, Basic set theory , Perspectives in mathematical logic, Springer-Verlag, Berlin, Heidel- berg, and New York, 1979
1979
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.